Perf batch 1

[timcadman]

Description of changes

• Refactored ds.abs, ds.asDataMatrix, ds.asInteger, ds.list, ds.asLogical, ds.asMatrix, ds.asNumeric.
• Call new serverside functions ds.exp and ds.log instead of the base function directly.
• Updated helper function .loadServersideObject to handle both object names and column names.
• Updated test expectations following refactor.

Instructions & checklist for PR reviewers

Code review

• Check any calls to exists and isDefined have been removed from clientside function, and appropriate checks added to server-side function.
• Check that client-side code checking whether an object has been successfully created has been removed
• Check whether any additional refactoring is required to reduce calls to serverside function.
• Check that function check relating to datashield connections object has been replaced with .set_datasources() (defined in utils.r)
• Check that performance tests have been added for all functions unless they already existed.
• Check tests have been updated to reflect changed expectations.

Testing

• Clone dsBaseClient using this branch.
• Install dsBase on a test server based on the branch perf-batch-1-v5.
• Run devtools::test( ) on the clientside package and check all tests pass
• Run devtools::build( ) on the clientside package and check it builds without errors
• Run devtools::test( ) on the serverside package and check all tests pass
• Run devtools::build( ) on the serverside package and check it builds without errors

[StuartWheater]

I am a bit concerned that the "tidying of the code", for example removing spaces, could cause problems when we come to merge v6.3.6-dev, v6.4.0-dev, ...., lots of conflicts. Could these changes be moved to a PR on v6.3.6-dev, moving them earlyed in release history.

I notices at the tests don't check that the results for the call which don't now return anything, I think that should check there are any results as intended.

[timcadman]

Thanks Stuart, this is still draft so I was going to work a bit more before asking for your review! I will revert any changes which change whitespace etc. I also need to check tests further.

[StuartWheater]

I also need to check tests further.
But if it did return a results wouldn't that be incorrect behaviour? So should be checked (I will generally put that in a "test-arg-..." test).

[timcadman]

It should return a result, but we can check that the serverside object has been created correctly? I will look today

[StuartWheater]

It would be an improvement if assignments could return a limited outcome.

[timcadman]

What outcome did you have in mind? A success message?

…

________________________________ From: Stuart Wheater ***@***.***> Sent: Friday, March 13, 2026 11:40 AM To: datashield/dsBaseClient ***@***.***> Cc: Cadman, TJ (medgen) ***@***.***>; Author ***@***.***> Subject: Re: [datashield/dsBaseClient] Perf batch 1 (PR #656) [https://avatars.githubusercontent.com/u/4047704?s=20&v=4]StuartWheater left a comment (datashield/dsBaseClient#656)<#656 (comment)> It would be an improvement if assignments could return a limited outcome. — Reply to this email directly, view it on GitHub<#656 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AJ4MXRMVV5PFXA2HYLGOOB34QPQTHAVCNFSM6AAAAACWQAPZ2GVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DANJUGE4DSOBSHA>. You are receiving this because you authored the thread.Message ID: ***@***.***>

________________________________ De inhoud van dit bericht is vertrouwelijk en alleen bestemd voor de geadresseerde(n). Anderen dan de geadresseerde(n) mogen geen gebruik maken van dit bericht, het niet openbaar maken of op enige wijze verspreiden of vermenigvuldigen. Het UMCG kan niet aansprakelijk gesteld worden voor een incomplete aankomst of vertraging van dit verzonden bericht. The contents of this message are confidential and only intended for the eyes of the addressee(s). Others than the addressee(s) are not allowed to use this message, to make it public or to distribute or multiply this message in any way. The UMCG cannot be held responsible for incomplete reception or delay of this transferred message.

[timcadman]

My thought process for this was that with R, you don't receive a message confirming that the process has worked. You just receive an error message. So I was thinking we do the same with DataSHIELD - let the server handle/return any errors, and if nothing is returned then it indicates the operation was successful